INFORMATION AND CONTROL SYSTEMS: ISSN 2664-9969
INFORMATION TECHNOLOGIES
UDC 004.42
DOI: 10.15587/2706-5448.2021.229532
Article type «Reports on Research Projects»
Kateryna Kyivska, IMPLEMENTATION OF ARTIFICIAL
Svitlana Tsiutsiura
INTELLIGENCE IN THE CONSTRUCTION
INDUSTRY AND ANALYSIS OF EXISTING
TECHNOLOGIES
The object of research is the process of using information technology in the construction industry. One of the
most problematic areas is increasing the efficiency of the construction industry through the introduction of digital
technologies. The research carried out is based on the application of an approach that is implemented using artifi
cial intelligence. The study used machine learning and fuzzy logic methods to mark visual data and analyze it for
potential threats, as well as to reduce all possible risks. The main feature of this approach is that using machine
learning technology, it is possible to reduce the risks of a project before they affect its profit. So, using artificial
intelligence in combination with BIM technologies, it is possible to predict work on construction projects based on
real-time data, past activities and other factors in such a way as to optimize construction processes. The benefits
to be gained from implementing digital processes will become even more evident in future projects as AI continues
to analyze company data. This is due to the fact that the proposed approach using fuzzy logic has a number of
features, in particular, the more information machine learning algorithms process, the more complex they become.
As a result, they provide even more useful information and allow to make even better decisions. This provides an
opportunity to minimize risks and efficiently allocate resources when working on projects. Compared to conventional
information technology, artificial intelligence can be used to build a knowledge-based security management system
and combine statistical probabilities to help mitigate security risks in construction projects.
Keywords: artificial intelligence, information technology, BIM technology, machine learning, automation of
the construction industry.
Received date: 14.01.2021 © The Author(s) 2021
Accepted date: 19.02.2021 This is an open access article
Published date: 30.04.2021 under the Creative Commons CC BY license
How to cite
Kyivska, K., Tsiutsiura, S. (2021). Implementation of artificial intelligence in the construction industry and analysis of existing technologies. Technology
Audit and Production Reserves, 2 (2 (58)), 12–15. doi: http://doi.org/10.15587/2706-5448.2021.229532
1. Introduction 2. Methods of research
The construction industry is currently undergoing a trans In order to find a way to close this gap, the following
formation from traditional, hierarchically organized construc two research questions were developed.
tion sites to digital and more autonomous ones. A digital 1) What are the potential benefits of implementing
shift is taking place and digital advancement is advancing artificial intelligence in the construction industry?
so rapidly that the industry is struggling to keep up. The 2) How is the construction industry benefiting from
introduction of artificial intelligence is a relatively unex the introduction of artificial intelligence today?
plored topic, especially in the construction industry [1]. This study is limited to case studies of a construction
Artificial intelligence (AI) can automate several operations project located in Oslo, Norway. The case study is limited
and improve the efficiency of the construction process [2]. to three digital tools. Based on the lack of published re
This is consistent with lean thinking, where minimizing search on the AI implementation, document examinations
waste and maximizing cost is the primary goal of AI adop and interviews were selected as specific methods.
tion. Therefore, it is relevant to analyze artificial intel The case in question is Bispevika, a construction project
ligence methods with the possibility of their combined in Oslo [3]. The project consists of approximately 7 years
use with BIM technologies. Thus, the object of research of construction and AI implementation in the construc
is the process of using information technology in the tion industry, and its impact on human behavior. This
construction industry. The aim of research is to show how project was chosen based on its unconventional way of
the construction industry can narrow the gap between working with innovative processes and tools. In addition,
the potential benefits and the up-to-date benefits from the choice was influenced by the fact that a project of
the AI implementation. this scale is an ideal testing ground for implementation
12 TECHNOLOGY AUDIT AND PRODUCTION RESERVES — № 2/2(58), 2021
Electronic copy available at: https://ssrn.com/abstract=3848553
ISSN 2664-9969 INFORMATION AND CONTROL SYSTEMS:
INFORMATION TECHNOLOGIES
through the possibility of training and improvement by the Digitization and digital transformation are often referred
same people. The Bispevika mentality is based on a lean to as the same thing. However, these words represent diffe
approach. To get an idea of the implementation of artificial rent levels of the digital world. Digitizing involves creating
intelligence, the implementation of three digital tools was a digital version of analog information, such as checklists,
considered in more detail (Table 1) [3]. on your mobile device instead of paper. Digitization refers
to the second level, where machines can perform human-
Table 1 controlled processes, such as the continuous updating of
Description of selected digital instruments at Bispevika the building information model (BIM technology) [6, 7].
Digitization Implementation
Digital transformation is the integration of digital technolo
Instrument Description gies in such a way that an organization is fundamentally
level phase
changing the way they work. AI is defined as using the
A web building collaboration
ability of a machine to mimic the intellectual behavior of
tool that can be thought of as Untimely
Touchplan Digitization a person, thus seeking to use human-developed algorithms
a digital version of the tools performance
in LPS (Sensory Plan 2018) for the approximate execution of conditionally complex
tasks. The introduction of artificial intelligence can be
4D digital construction platform
empowers workers to visua seen as a digital transformation. The main benefits of AI
Intermediate for organizations are: improving the end-user experience
Synchro lize, discuss and collaborate Digitization
implementation
to find all constraints before with the data it provides [8]. Automating tasks allows
doing (Synchro 2019) people to focus on work that will add value and reduce
Artificial intelligence scheduling human error, and deliver services faster. This can be seen
software that uses user input to in connection with the principles of lean manufacturing,
Digital
ALICE create and optimize schedules Early testing where maximizing value, minimizing waste and improv
conversion
based on the created methods ing the efficiency of the work process are central [9].
(ALICE 2017)
As is well known, AI adaptation is relatively low in the
construction industry. However, there are several startups
In order to identify useful experience in the AI imple today that offer programs related to scheduling and image
mentation in other industries, a number of studies were recognition. Using historical data in addition to human
carried out, which were based on identifying the digital resources, algorithms can look at millions of alternatives
shift and its impact on human behavior [4]. for projects and continually improve graphs. Image re
cognition allows hazardous workers to be identified and
3. Research results and discussion aggregated to determine future training and education
priorities. However, any AI algorithm is based on learning,
There is a perception that the construction industry is not programming, which means that algorithms require
digitalized and has a low level of productivity development a certain amount of data to be executed at the human
compared to other industries. However, research shows level. Getting big data is seen today as a limitation for
that the correct implementation of digitization will in many construction companies. Artificial intelligence can
crease the efficiency of the construction process. More help the construction industry overcome the industry’s
advanced technologies such as artificial intelligence are biggest challenges, including costs, planning and safety.
now entering the industry [5]. However, the construction industry is now in the early
With a more or less complete lack of research looking stages of digital transformation, and few projects have
at the AI implementation, it will be necessary to take actually implemented artificial intelligence. The central
advantage of the research results concerning the imple question that is considering implementation will be how
mentation process itself and combine them with the AI to achieve successful human-AI collaboration.
technical knowledge. Document studies and interviews were The digital shift introduces new ways of working, where
selected as appropriate data collection methods (Table 2). people and technology must collaborate. Traditional cul
ture classifies people and technology as separate entities.
Table 2 However, in recent years, technological development has
Description of external interviews led to the emergence of advanced automation that can
respond better than humans in specific situations. Humans
Company Company description Interview description
can’t compete with AI when it comes to analyzing data,
A research center that information and knowledge, just as AI can’t compete with
Norwegian
Оpen AI
brings together various Professor of Computer Science and human abilities for pedagogy, creativity, vision and ethics.
research efforts within the Artificial Intelligence Since successful cooperation between humans and humans
lab
AI framework
requires certain tasks and responsibilities, then coopera
Currently, the key research Extensive experience in the de tion between humans and artificial intelligence requires
areas are healthcare, ener velopment and implementation of the same. Although tasks and responsibilities are split bet
Inmeta gy, ocean space, digital artificial intelligence systems in the ween human and AI, it can be difficult for humans to
economy and smart envi oil and gas industry, gas industry trust the AI output. How does the AI arrive at a certain
ronment (NTNU 2019) and fishing
prediction and recommendation, and then determine if
Data scientist with a specialization this is a good decision or not? This is one of the rea
Space
One of the leading machine in machine learning. Experience in sons that the adoption of some AI tools remains low in
learning consulting compa the development and implementa application areas where clarification is useful or really ne
maker
nies in Norway (in 2019) tion of artificial intelligence systems
cessary. Trust includes a complex set of interacting factors,
in the healthcare sector
including attitudes, beliefs, control, emotions, risk, and
TECHNOLOGY AUDIT AND PRODUCTION RESERVES — № 2/2(58), 2021 13
Electronic copy available at: https://ssrn.com/abstract=3848553
INFORMATION AND CONTROL SYSTEMS: ISSN 2664-9969
INFORMATION TECHNOLOGIES
power. So the AI has no attitudes, beliefs or emotions, this Table 3
can lead to the fact that it will be difficult for a person Future applications of artificial intelligence presented
to trust the result. The person is in a vulnerable posi by the Bispevika team
tion. However, if vulnerability is defined as ability (op
AI category Application Influence
portunity, intention), and it is assumed that the inten
tion of the AI coincides with the intention of the person, Analysis of large amounts of data based
then people can study the capabilities of the AI through Machine
on historical figures and human con
training. In other words, transparency and trust can be Planning tributions. Evaluates millions of layouts
learning
that take exponentially more time for
increased by teaching people what AI can and can’t do, people to complete
and where violation can occur. However, what happens if
humans and artificial intelligence disagree? Fluid interac Health, safety and
Prediction/early detection of dange
tion, which allows the person to edit the AI, and the AI Pattern rous situations using machine learn
labor protection,
recognition ing algorithms combined with pattern
to guide the person, will be imperative. The goal is to environment
recognition
enable collaborative data exploration, leading to a com
mon framework where AI and human beliefs have been Digital map that constantly shows where
updated. The interaction that allows AI to answer such Pattern and where materials or machines can
Warehouse space
recognition be stored. Improves predictability and
questions as «why do I need to change my plan at this efficiency in the field
time?» will increase a person’s ability to find biased data
and understand and trust AI. Another important factor in Jobs doing
For example, work at height. Fire pro
trust is time. If the machine runs reliably and predictably Automation tection of steel beams. Increase the
hazardous work
safety of people
for a long time, people will begin to trust AI as much
as they trust other people. Self-propelled The use of robots and self-propelled
The follow-up and findings are based on three external Automation construction equip construction equipment will change the
interviews with a high level of AI knowledge and nine ment workflow
case studies with a relatively low level of AI knowledge. The jobs that travel around the con
The remaining eight respondents on specific cases are not struction site scan the situation on site
included in the study due to their lack of knowledge Automation Quality assurance and compare it to the BIM project. This
about AI. The interviewees were asked an open-ended technology can save hundreds of hours
spent on quality
question that looks at the future of artificial intelligence
in the construction industry to reflect people’s thoughts on
the digital future of construction. This question led to both Touchplan and Synchro are transparent technology in
examples of use and possible consequences. In analyzing which a person controls how an input signal is converted
the collected data, applications were divided into three into an output signal. The result of a human-driven process
categories of AI: machine learning, pattern recognition, and that makes the process transparent means a simple and
automation. Machine learning [10] consists of algorithms reliable result. However, there is no guarantee that it is
that can analyze data and develop new solutions based the most optimized, for example, in terms of progress,
on historical figures. Pattern recognition is the process of staffing and cost, since the human brain has a limited
pattern recognition using a machine learning algorithm. ability to see possibilities.
Automation is the process by which AI will be able to ALICE is an artificial intelligence-based construction
perform tasks that are first performed by humans, or tasks platform that uses AI to analyze the flow of labor and
that are beyond human capabilities [11]. Table 3 presents equipment across the entire jobsite. This technology allows
the structure of the polls for these three categories [4]. users to explore different construction plans in real time
All AI categories in Table 3 will affect the workflow and understand the impact of key building decisions on
and work tasks. The industry today puts trust in individual the cost and duration of a project.
expertise over empiricism.
However, to prove the possible benefits of artificial intel 4. Conclusions
ligence in the construction industry, a study was conducted
on the implementation of three digital tools: Touchplan, The study confirmed the high potential of using artifi
Synchro and ALICE [12]. cial intelligence and machine learning in the construction
Touchplan is a construction planning tool that takes industry. The implementation of the proposed approaches
this technology collaboration to the next level. The main will improve the efficiency of construction, its safety and
goal of the tool is to create exceptional teams that deliver quality. To realize the potential of big data in construc
impressive projects – ones that run smoothly, complete on tion through machine learning and artificial intelligence
time or ahead of schedule, and use the collective intel algorithms, it is necessary to accumulate a critical mass of
ligence of the entire team to learn and solve problems data on various construction projects and integrate vari
every day. ous information systems of construction companies. The
Synchro is a 4D digital construction platform that gives results of the study will be useful in the standardization
workers the ability to visualize, discuss and collaborate of construction processes and the processing of data on
to find all constraints before completing projects. As with construction projects.
Touchplan, there is currently no specific 4D planning train In the course of further research, it is planned to con
ing strategy or training in Synchro. The main advantage is sider the introduction of artificial intelligence, taking into
the use of the 4D model to gain a general understanding account the needs of users, based on the collection of
of work tasks. This model is especially useful in terms of empirical data and the study of experience gained from
the diversity of languages at the construction site. other industries.
14 TECHNOLOGY AUDIT AND PRODUCTION RESERVES — № 2/2(58), 2021
Electronic copy available at: https://ssrn.com/abstract=3848553
ISSN 2664-9969 INFORMATION AND CONTROL SYSTEMS:
INFORMATION TECHNOLOGIES
References 8. Mikhailenko, V. M., Terentiev, O. O., Shabala, Ye. Ye., Kyiv
ska, K. I., Horbatiuk, Ye. V. (2017). Modeli, metody ta infor
1. Casebeer, W. D.; Masakowski, Y. R. (Ed.) (2020). Building matsiina tekhnolohiia diahnostyky tekhnichnoho stanu budivelnykh
an Artificial Conscience: Prospects for Morally Autonomous konstruktsii i sporud. Kyiv: TsP «Komprynt», 161.
Artificial Intelligence. Artificial Intelligence and Global Security. 9. Terentyev, O., Bohdan, M. (2015). The Method of Prediction
Bingley: Emerald Publishing Limited, 81–94. doi: http://doi.org/ of Deformations of Buildings and FІlure Analysis the Examina
10.1108/978-1-78973-811-720201005 tion of Technical Condition of Buildings. International Journal
2. Sacha, D., Sedlmair, M., Zhang, L., Lee, J. A., Peltonen, J., of Science and Research, 4 (8), 280–282.
Weiskopf, D., Keim, D. A. et. al. (2017). What you see is what 10. Terentyev, О., Poltorak, О. (2017). Risk assessment of delayed
you can change: human-centered machine learning by interac damage diagnostics of technical condition of building structures.
tive visualization. Neurocomputing, 268, 164–175. doi: http:// ScienceRise, 2 (31), 42–45. doi: http://doi.org/10.15587/2313-
doi.org/10.1016/j.neucom.2017.01.105 8416.2017.93907
3. Hunn, L. K., Fyhn, H.; Pasquire, C., Hamzeh, F. R. (Ed.) (2019). 11. Terentyev, О., Poltorak, О. (2016). Development of models and
Building and Sustaining a Culture with a Mindset for Disruptive methods for determining the physical deterioration of items for
Performance: A Case Study from Bispevika Norway. Proc. 27-th the task of diagnostics of technical condition of buildings and
Annual Conference of the International. Dublin: Group for Lean Con structures. ScienceRise, 8 (2 (25)), 14–19. doi: http://doi.org/
struction (IGLC), 369–378. doi: http://doi.org/10.24928/2019/0172 10.15587/2313-8416.2016.76318
4. Walday, M., Olsgard, F. (2004). Ny senketunnel I Bj rvika. Biolo 12. Abbass, H. A. (2019). Social Integration of Artificial Intelligence:
giske forunders kelser i November 2003. Rapport 0-VK-203, 30. Functions, Automation Allocation Logic and Human-Autonomy
5. Terentiev, O. O., Balina, O. I., Shabala, Ye. Ye., Turushev, O. S. Trust. Cognitive Computation, 11 (2), 159–171. doi: http://
(2016). Model definition of physical deterioration of structural doi.org/10.1007/s12559-018-9619-0
elements the building for the tasks of diagnostics of technical
condition. Management of Development of Complex Systems, 26,
153–157 Kateryna Kyivska, PhD, Associate Professor, Department of Infor
6. Kyivska, K. I., Tsiutsiura, S. V., Tsiutsiura, M. I., Kryvoruchko, O. V., mation Technology, Kyiv National University of Construction and
Yerukaiev, A. V., Hots, V. V. (2019). A study of the concept of Architecture, Kyiv, Ukraine, ORCID: https://orcid.org/0000-0003-
parametric modeling of construction objects. International Journal 0906-1128, e-mail:
[email protected] of Advanced Research in Engineering and Technology, 10 (2),
636–646. doi: http://doi.org/10.34218/ijaret.10.2.2019.060
Svitlana Tsiutsiura, Doctor of Technical Sciences, Professor, Depart
7. Terentyev, O., Tsiutsiura, M. (2015). The Method of Direct
Grading and the Generalized Method of Assessment of Build ment of Information Technology, Kyiv National University of Con
ings Technical Condition. International Journal of Science and struction and Architecture, Kyiv, Ukraine, ORCID: https://orcid.org/
Research, 4 (7), 827–829. 0000-0002-4270-7405, e-mail:
[email protected]TECHNOLOGY AUDIT AND PRODUCTION RESERVES — № 2/2(58), 2021 15
Electronic copy available at: https://ssrn.com/abstract=3848553
